network-accessibility-analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The required workflow described in SKILL.md ingests/uses external network and GTFS sources (public OSM and GTFS feeds) at runtime to construct routing graphs and validate calendars, which are outsider-authored free-text fields and can carry indirect prompt-injection payloads into the agent’s LLM context via the retrieved feed/page content or extracted strings.